Merge branch 'master' into xsn/common_json

common/arg.cpp

@@ -1897,7 +1897,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         [](common_params & params, bool value) {
             params.conversation_mode = value ? COMMON_CONVERSATION_MODE_ENABLED : COMMON_CONVERSATION_MODE_DISABLED;
         }
-    ).set_examples({LLAMA_EXAMPLE_COMPLETION, LLAMA_EXAMPLE_CLI}));
+    ).set_examples({LLAMA_EXAMPLE_COMPLETION}));
     add_opt(common_arg(
         {"-st", "--single-turn"},
         "run conversation for a single turn only, then exit when done\n"

common/speculative.cpp

@@ -2322,6 +2322,9 @@ common_params common_base_params_to_speculative(const common_params & params) {
     const auto & params_spec = params.speculative.draft;
     common_params result = params;
 
+    result.embedding    = false;
+    result.pooling_type = LLAMA_POOLING_TYPE_UNSPECIFIED;
+
     if (has_draft) {
         result.devices               = params_spec.devices;
         result.model                 = params_spec.mparams;